DE10147484A1 - Control of the injection cycle of a fuel injection engine, involves writing back control cycle operating parameters to control memory so that control computer can monitor control process and make adjustments if necessary - Google Patents

Abstract

A method for controlling actuators, especially fuel injectors of a fuel injection engine. A central engine control computer unit (1) writes initial control parameters to a control memory (3) or parameter RAM of a gate array. The control cycle is measured during operation and measurement values relating to the control cycle are written back to the corresponding control memory. The invention also relates to a corresponding fuel injector control arrangement, computer program and computer program product.

Description

The invention relates to a method for controlling Actuators, in particular from actuators for Fuel injection with an injection engine, at which method of control is necessary Control parameters from a central processing unit in entered a memory assigned to the control be, as well as a corresponding facility for Implementation of such a procedure. The invention still relates to a corresponding computer program as well as computer program product.

State of the art

With today's control concepts for inductive actuators (for example solenoid valves of injection engines) from a central processing unit (CPU) values for the Actuation history (start and duration or end of actuation) in Control signals implemented for the power amplifier module. to Relief of the CPU are often used in gate arrays for (injection) -synchronous control of the most (Injection) system-specific power amplifier modules are used. The ones written in time by the CPU to the gate array Control parameters are via the state machine, the timer etc. of the gate array in the control signals for the Power stage module converted.

However, there are several disadvantages with this concept. To the one is remembered by the central processing unit (CPU) in the further program flow not which Control parameter values were sent to the gate array. With the concepts mentioned, each value is also a fixed one Assigned address in the parameter RAM of the gate array. This is runtime, but not optimal memory. Finally the central processing unit has no possibility from which Correcting knowledge of the real control process or to regulate the actuators. This is only possible due to the system reactions of the engine.

In the present invention, the above Disadvantages are overcome and the control process at Fuel injection optimized for storage and runtime will be realized.

Advantages of the invention

According to the control curve of an actuator are recorded and measured values of this control curve are recorded in Areas of the memory assigned to the control written back. This makes a bidirectional Data flow possible and necessary, so that from the central Computing unit intervenes in the control process can be. The corresponding facility contains funds for recording the control curve of an actuator and for Transmission of the measured values in certain memory areas of the memory assigned to the injection process. This Memory is, for example, the one mentioned at the beginning Parameter RAM of a gate array or one of its tasks adopting TPU (Time Processing Unit).

According to the invention, the measured values of the control curve of the actuator to the, especially dynamically allocated Entries of the control parameters in the allocated memory (Parameter RAN) written back (from the TPU to the CPU).

It is advantageous if writing back the measurement values of the control process only takes place when the relevant control process is completely completed. The completion of a control process can, for example can be coded by setting a bit. To avoid It therefore makes sense to overlap those to be controlled Actuators in different blocks or banks like this to arrange that successive control processes do not overlap within a block or bank.

It also makes sense not only to write back the Measured values, but also the entry of the control parameters to be dynamic, with an entry only then takes place when the central processing unit in the to descriptive memory area Read measured values of the control process.

Several values come in as measured values of the control curve Question. In particular, the activation duration and the ID of the respective actuator can be returned. any Errors can be encoded and returned as error status become. Especially when triggering Piezo actuators, such as in fuel injection, can as the characteristic values characterizing the control process, the charging, Hold and discharge time of the piezo actuator as well as the reached Actual voltage can be written back.

The central Computing unit at least the start of control and the Transfer control duration (or the control end). Further Parameters can be the type of control process, for example the type of injection (pre-injection, main injection or post-injection) Actuator and the setpoints for example for the Drive voltage.

An inventive device for controlling Actuators has a central processing unit, one of the Control assigned memory and means for Enter the necessary for control Control parameters from the central processing unit in the the control assigned memory, with means for Detection of the control curve of an actuator and Transmission of the measured values in certain areas of this Control associated memory are provided.

The device for performing the invention The method advantageously has the central one Computing unit and one for the control process responsible computer on a parallel bidirectional interface are interconnected. As the computer responsible for the activation process serve a TPU that does the job of the aforementioned Gate arrays can take over. The CPU stands with in particular the parameter RAM of the TPU. The structure and exact functioning of such a facility is in Connection with the embodiment below explained.

A computer program according to the invention controls and processes the data flow from the central processing unit (CPU) and the memory assigned to the control process (Parameter RAM of the TPU) to the inventive method to be able to execute when loading the computer program. The Computer program can from the central processing unit, but also from an external computing unit (Injection computer) can be started. The program code Means of the computer program are on one computer readable disk stored in the system or are downloaded from an external source. As Data carriers come EEPROMs, flash memories, but also Hard drives, floppy disks or CD-ROMs in question.

It should be noted that the process and the Device for controlling actuators in the above described form generally for event-driven Applications are suitable. Such event-driven Applications are, for example, fuel injection with an injection engine or the braking process with a Motor vehicle. But also outside of Automotive applications are applications of inventive method and the inventive Setup makes sense wherever the control of a process necessary control parameters from one central processing unit in one of the control allocated memory are transferred. As a possible The area of application is fuel injection in an injection engine, so in the following for reasons the better understanding of the use of the invention limited to this area.

The invention is intended to be based on a attached figures illustrated embodiment are explained in more detail.

drawings

Fig. 1 shows schematically a control and memory unit for controlling actuators according to the invention for fuel injection in an injection engine.

Simplified Fig. 2 shows the driving operation in a 6-cylinder fuel injection engine with pilot, main and post injection.

Fig. 3 shows schematically the inventive dynamic entry distribution of the control parameter and the measured values of the back written Ansteuerverlaufs.

The following embodiment relates to a 6- Cylinder injection engine, being used as actuators that Control fuel injection, use piezo actuators come. It should be noted that the invention also for other injection engines and other actuators (inductive Actuators such as solenoid valves) can be used. The the following explanations are therefore in no way to be understood restrictively.

Fig. 1 shows a device for actuator control of the present invention in simplified schematic form. The central processing unit (CPU) 1 is connected via an internal bus 2 to the module 7 for controlling the actuators of the injection engine. The data from the internal bus 2 to the TPU (Time Processing Unit) 4 and to the parameter RAM 3 are received via the control functions 5 . In principle, the TPU 4 , like the gate array of previous computer concepts, serves to control and diagnose the power amplifier modules that directly control the actuators. For this purpose, the TPU 4 is connected to other components via the local bus 6 . The parameter RAM 3 serves as memory for the two TPU units 4 . The entire module 7 forms a parallel bidirectional interface between the CPU 1 and the power amplifier modules (dual port TPU RAM, DPTRAM).

The assignment of parameter RAM 3 is basically divided into two banks. This ensures that possible system-related overlaps of actuator controls are avoided. Fig. 2 shows an example of the distribution of Aktoransteuerungen for a 6-cylinder engine with the firing sequence 1, 2, 3, 4, 5, 6, wherein of the bench 1, the cylinders 1, 3 and 5, from the bank 2, the cylinder 2 , 4 and 6 can be controlled. In the example shown in FIG. 2, four different types of injection are shown, namely two pre-injections and one main and one post-injection each. The figure shows a section of two crankshaft revolutions (720 °) in which all cylinders are active once. Two consecutive bars are shown for each injection type, the first bar covering the period of time that the respective injection can begin, while the second bar indicates the longest possible duration of this injection. In other words, the second bar can be moved within the first bar. From this it can be seen that although there may be overlaps between two cylinders (cylinders 1 and 2 ) to be controlled in succession, overlaps between the control processes of the cylinders of a bank (cylinders 1 and 3 ) are rather rare.

In the following, only the entry distribution for one of the two banks with two different injection types (Pre and main injection) are considered.

For software timing: In static interrupts, the start of actuation of the respective injection type, the injection type, the voltage level (for the piezo actuator) and possibly other parameters are calculated in the CPU 1 and written to the parameter RAM 3 . In the dynamic interrupt, which takes place before the start event, the activation duration (or the activation end) is calculated in the CPU 1 and written into the parameter RAM 3 . (The reason for this is that in common rail systems the control duration is a function of the rail pressure, so that the measurement takes place before the injection due to the strongly fluctuating rail pressure.) Before the values are entered into the parameter RAM 3 by the CPU 1 the actually measured activation duration (or the activation end) as well as further values such as injection type, voltage level reached, etc., of an injection in the past in the static and / or dynamic interrupt (of CPU 1 ).

It is not necessary to reserve entries and thus storage space for each actuator of the respective bank, since the injections from actuators of a bank usually do not overlap in practice. Since the assignment of the parameter RAN is not permanently assigned to specific actuator numbers (e.g. solely through the address information), the actuator number information is also stored in the entry. Once the control parameters have been entered, the events are processed by the TPU 4 .

The TPU 4 measures the control curve, e.g. B. the charging, holding and discharging time of the piezo actuator, and writes the values back into the memory subareas for the control parameters. Before the CPU 1 rewrites the parameter frame area, it must first read and save the return values.

Fig. 3 shows the entry distribution in the concrete exemplary embodiment of a three cylinders (cylinders 1, 3 and 5) which drives bank.

For the sake of clarity, as I said, only the preliminary and main injection of the cylinders. Between two static interrups on one bank can no longer programmed as three injections (or on processing) his. This results from the customer or System requirements. Every entry is with AB for Start of activation and AD for activation duration. in the The example shown is this bank division dissolved, d. H. there are a total of six possible entries. The bank information is entered as a parameter by the CPU in entered the respective entry of the parameter RAM. This gives you more flexibility when z. B. on one Bank between two static interrupts four Injections programmed (or are in progress) and on the other bank two or less than two Injections expire.

The first three lines of the table shown in FIG. 3 represent the chronological order of these entries. The fifth line gives the chronological order with n - 4, n - 3,. , .n,. , ., n + 3 again, while the fourth line denotes the respective injection, identified by the injection type and the actuator number (VE stands for pre-injection, HE stands for main injection, Z1 to Z5 stands for cylinder 1 to cylinder 5 ). The last line shows the measurement result of the injection, which is written back at the given time (fifth line), whereby in this last line, in addition to the injection type (VE, HE) and the actuator number (Z1, Z3, Z5), the associated time is also given that belongs to the measurement result.

Basically, it must be ensured that the Return values can be picked up in time. Would between two static interrupts only one injection type (e.g. HE) programmed, so it would be enough, always only the first Read out and describe the entry. But that would have been Consequence that of the other two entries Return values of past injections are not would be read out. To prevent this, a Entry pointer created according to the ring buffer principle (here three entries per bank) with each one to be written Entry moves an entry field in the parameter RAM. This ensures that the return values are already processed entries at all and on time be read out. If the entry is not yet free, i. H. the event has not been processed, he moves Entry pointer one entry further.

In the example in FIG. 3, the control parameters for the main injection of the cylinder 3 are written into the parameter RAM at time n-1 (see first line). At this point, the measured values for the pre-injection of cylinder 1 must be read and saved by the CPU (see last line in the table). Otherwise, as can be seen from line 1 of FIG. 3, these measured values would be overwritten without being returned to the CPU. After the main injection of cylinder 3 has been completed , the measured values of this injection process are written back as entry 1 at time n and are then read back at time n + 2 at the latest. At said time n, the control parameters for the pre-injection of the cylinder 5 are written into the parameter RAM as entry 2 by the CPU. The measurement results for the pre-injection of the cylinder 3 must have been read back by the following time n + 1 at the latest, since the control parameters for the main injection of the cylinder 5 are then given as entry 3 by the CPU to the parameter RAM. At time n + 2, the entry for the pre-injection of cylinder 1 begins (first line). At this time, the measurement results for the main injection of cylinder 3 must have been read and saved by the CPU. This process runs continuously according to the ring buffer principle.

By increasing the number of entries or by the above The described breakdown of the bank structure can be further Realize injection types. Basically, the following applies Execution of an injection: in that the Activation start and activation duration (or activation end) Events are processed, these entries (RAM Cells) with measured values, such as real control duration, charging, Discharge times, etc., so that a Multiple use is possible.

By means of the invention, the behavior of the monitor individual actuators, errors can be detected recognize that may also directly from the TPU the CPU can be reported. The invention thus replaces unidirectional control of the injection process a bidirectional control of the same with the hereby related benefits.

Claims (9)

1. A method for actuating actuators, in particular actuators for fuel injection in an injection engine, in which actuation parameters necessary for actuation are entered by a central processing unit ( 1 ) into a memory ( 3 ) assigned to the actuation, characterized in that the actuation curve of an actuator recorded and measured values of the control curve are written back into areas of the memory ( 3 ) assigned to the control. 2. The method according to claim 1, characterized in that that the measured values are only written back, if the relevant control process is complete is completed. 3. The method according to claim 1 or 2, characterized in that the input of the control parameters into the memory ( 3 ) takes place dynamically, in particular only when the central processing unit ( 1 ) has read out the measurement values of the control process written back into this memory area ( 3 ) , 4. The method according to any one of claims 1 to 3, characterized characterized that as a measured value of the control curve or more of the following values, control duration and any error data and, especially when using Piezo actuators, the actual voltage, charging, holding and Discharge time, as well as an actuator identifier, written back become. 5. The method according to any one of claims 1 to 4, characterized characterized that as the control parameter of Start of activation, duration of activation or end of activation as well as other parameters such as type of control process, actuator and setpoints are used. 6. Device for controlling actuators, in particular actuators for fuel injection in an injection engine, with a central processing unit ( 1 ), a memory ( 3 ) assigned to the control and with means for entering control parameters required for control from the central processing unit ( 1 ) in The memory ( 3 ) assigned to the control, characterized in that means are provided for detecting the control curve of an actuator and for transmitting the measured values to specific areas of this memory ( 3 ) assigned to the control. 7. The device as claimed in claim 6, in which the central processing unit ( 1 ) is connected via a parallel bidirectional interface to a processing unit ( 4 ) responsible for the control process. 8. Computer program with program code means for performing a method according to one of claims 1 to 5, if the computer program on a computer, in particular on the central processing unit ( 1 ) and / or the processing unit responsible for injection ( 4 ) according to claim 6 or 7, is running. 9. Computer program product with program code means which are stored on a computer-readable data carrier in order to carry out the method according to at least one of claims 1 to 5 if the computer program product on a computer of a corresponding computing unit, in particular the central computing unit ( 1 ). and / or the computing unit ( 4 ) responsible for the injection according to claim 6 or 7.